The broad, long-term objectives are (1) demonstrate computational and experimental methods cooperating to achieve a functional census of a large mutation sequence space of great medical importance; (2) contribute to our knowledge of p53 functional rescue mechanisms, and so facilitate the search for a small molecule cancer drug that effects an analogous functional rescue of p53; and (3) elucidate part of the systems biology of cancer by characterizing the spectrum of p53 cancer and suppressor mutants across known downstream p53 DNA binding sites. Mutations to the tumor suppressor protein p53 occur in approximately half of all human cancers, and restoring function to a mutationally defective p53 protein is a long-held medical goal. Biological precedence for rescuing p53 cancer mutations is found in second-site p53 cancer suppressor mutations. The analogous p53 pharmacological rescue would save hundreds of thousands of lives annually. Understanding and predicting p53 rescue is an important step toward that goal. The specific aims are (1) computationally predict all single suppressor mutations for p53 cancer mutants and validate the results experimentally, (2) optimize the rescue effects of known and putative p53 suppressor regions through two or more coordinated mutation changes, and (3) predict and experimentally validate the DNA binding specificity of p53 cancer and suppressor mutants for known p53 DNA binding sites. Our broad strategy is a coordinated computational and experimental attack. We already have experimental p53 functional assays and computational predictors of p53 activity, developed as part of our Preliminary Studies. Computational predictors will be used to focus experimental work into the highest priority areas. Experimental validation of the predictions will lead to a larger training set for machine learning techniques. The larger training set will lead to even more accurate predictions, thus even more focused experimentation. Thus, the interplay between computation and experiment will become ever more efficient as the project progresses. Variations of this basic strategy apply to each of our Specific Aims, which all rely on closely coordinated experiment and computation.